Enrichment/Depletion Strategies for Protein Analysis

Protocols that enrich samples in target or product proteins, while depleting unwanted protein interferences, form the foundation of sample preparation for proteomics and quality control for protein-based therapeutics, diagnostics, and reagents. Enrichment/depletion are therefore core activities in protein sample prep.

Thermo Fisher Scientific offers a five-item list of considerations for general protein sample prep:

Extraction involves mechanical, chemical, or enzymatic release of protein from its matrix and its subsequent solubilization separate from lipid and small molecule fractions. Choice of extraction method depends on the sample type and location of the protein inside the cell (e.g., inside specific organelles).

Preservation includes low temperature and the introduction of buffers that retain the target protein’s chemical integrity and activity.

Cleanup involves specific concentration and depletion steps, including dialysis, desalting, concentration, and removal of small molecule contaminants.

Quantification establishes the protein’s titer and allows back-calculating its original abundance in the sample.

Detection and measurement involve the application of specific analytical methods for more precise quantitation and further characterization.

Given the myriad sample sources, analysis endpoints, required limits of quantitation/detection, instrumental approaches, degree of regulatory adherence, and overall objectives calculating the number of potential workflows is impossible. It is safe to say, then, that every situation is unique.

The following sections examine a chromatography-based approach to purifying proteins from complex mixtures derived from human embryonic kidney (HEK) cells, which are commonly used to produce human therapeutics and diagnostics.

A look at a HEK293 prep

Conditioned medium samples from HEK293 cell cultures are extremely heterogenous, with varying numbers and quantities of proteins, including the “target” protein and host cell proteins (HCPs).

Typically, the target makes up less than 1% of total protein, at a concentration of less than 0.001 mg/mL. Its purification through standard methods to >95% purity and a concentration of 1mg/mL—standards with relevance to biopharmaceutical development as general good analytical practices—is time- and resource-consuming. Sample preparation has been a substantial bottleneck in most proteomic analysis workflows.

Scientists at Proteintech have developed and optimized an approach to target protein-specific purification based on FPLC (fast protein liquid chromatography)—a method normally used for both analysis and preparative work. Purification occurs through a series of steps, using stationary-phase resins selected for separations based on molecular weight, binding affinity, or chemical properties, resulting in a target protein purity greater than 95%.

“In addition to information on the general physical and chemical characteristics of the target protein, it is extremely useful to know as much as possible about the contaminating endogenous proteins in the sample, with the goal of eliminating them,” says Bulbul Pandit, Ph.D., Sr. Scientist at Proteintech.

The purification scheme involves sequential elution through two or more columns containing the appropriate resins. Each column purifies the protein of interest until the desired purity is reached.

“Ideally, this can be done using only one resin. However, on average, two different sequential column runs are required to achieve >95% purity,” Pandit tells Biocompare.

The secret to this method is the judicious use of aqueous buffer delivered at constant flow, but with the ability to alter the buffer composition, for example using different salts or gradients.

“Based on properties of the protein of interest, for example its charge, size, or hydrophobic characteristics, the selected resins, and the mobile-phase composition, we routinely achieve 95% homogeneity for target proteins,” Pandit adds.

The knock against chromatography in these situations is it is almost always dilutive. According to Pandit, this has not been an issue. “In many cases we can elute the protein from a chromatography column in a smaller volume than the volume in which it was originally loaded, thereby concentrating the protein. For example, if one liter of conditioned media is loaded onto the column and the protein elutes in 100 mL, we have effectively concentrated the protein tenfold. Some protein is of course lost during purification. However even with losses of 50% we have still concentrated the protein fivefold. We always have the option of adding a concentration step at the end to reach the desired protein concentration.”

The stationary phases consist of ion exchange, affinity, hydrophobic interaction, or gel filtration in standard crosslinked agarose formats and packed into cylindrical columns. Depending on the sample, the beads interact mainly with either the protein of interest or the contaminants.

Calculating relevant isoelectric points

“One important parameter to identify before developing a purification scheme is the isoelectric point (pI) of the target protein, which may be estimated from the pKa values for ionizable sidechain groups, and which may be used to optimize the buffer pH in subsequent ion-exchange steps,” Pandit says. Rather than worrying about the pIs of all known species, Proteintech focuses only on the value for the target.

This method works quite well but only makes sense when the cell line and the proteins it generates endogenously are known. Luckily this is the case in almost all production settings, e.g., CHO in mAb production. “We use HEK293 to express recombinant proteins so we are very familiar with most of the HCPs this line generates,” Pandit adds. “For example, almost 90% of endogenous host cell proteins will not bind to cation exchange resins at pH 7.4.”

It also helps to understand the properties of the target protein, particularly for the affinity chromatography step(s). Affinity chromatography for immunoglobulins is covered by protein A and similar molecules with “innate” attraction to those targets. Another approach involves the use of tags such as poly-histidine or glutathione S-transferase, which are attached to the protein, used in the purification step, and then usually removed.

Proteintech produces its cytokines and growth factors free of tags, so the company rarely uses the tagging-affinity approach.

Pandit explains that while tags are quite popular with IgGs, that method often leaves a few extra amino acids dangling from the product after the tag removal step, either at the N-terminus or C-terminus. “These extra amino acids may trigger a dangerous immune response in patients. It is possible to add a protein cleavage site between the tag and peptide, which allows for tag removal, but this opens up the possibility of cleaving the mature protein downstream. We take pride in offering authentic human proteins completely without tags.”

This article examined a tiny slice of the many workflow possibilities for protein enrichment and depletion as a front-end preparatory to instrumental analysis. Remember that while many protocols offer a more streamlined approach, including kit-based methods, using these methods always comes at a cost in data quality.